Pedestrians’s trust in AVs and whether these vehicles enable them to cross the street safely and efficiently are crucial factors that directly influence society’s acceptance of autonomous vehicles (AVs). External Human-Machine Interface (eHMI) have been evaluated as effective tools for facilitating interaction between AVs and pedestrians. In this study, we designed a VR experiment to explore the issue of eHMI position selection under different display timings. The experiment recorded participants’ (N = 30) subjective trust levels and crossing behavior during their interaction with the AV. The experiment manipulated two factors: the display timing of the eHMI (early display vs. late display) and the position of the eHMI (roof, windshield, and bumper). The results indicated that when the eHMI was displayed early, the roof and bumper positions were more effective, while no significant differences were found among the three positions when the eHMI was displayed late. These findings validate our hypothesis that the effect of eHMI position varies depending on the display timing. Overall, our results provide insights into the design of eHMI display timing and position for AVs and lay a foundation for future research on the dynamic processes of eHMI display.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Exploring the Effects of eHMI Position Under Different Display Timings on Pedestrian Trust and Crossing Behavior in VR

  • Lijun Jiang,
  • Lili Deng,
  • Zhelin Li,
  • Siu Shing Man

摘要

Pedestrians’s trust in AVs and whether these vehicles enable them to cross the street safely and efficiently are crucial factors that directly influence society’s acceptance of autonomous vehicles (AVs). External Human-Machine Interface (eHMI) have been evaluated as effective tools for facilitating interaction between AVs and pedestrians. In this study, we designed a VR experiment to explore the issue of eHMI position selection under different display timings. The experiment recorded participants’ (N = 30) subjective trust levels and crossing behavior during their interaction with the AV. The experiment manipulated two factors: the display timing of the eHMI (early display vs. late display) and the position of the eHMI (roof, windshield, and bumper). The results indicated that when the eHMI was displayed early, the roof and bumper positions were more effective, while no significant differences were found among the three positions when the eHMI was displayed late. These findings validate our hypothesis that the effect of eHMI position varies depending on the display timing. Overall, our results provide insights into the design of eHMI display timing and position for AVs and lay a foundation for future research on the dynamic processes of eHMI display.